[mw] Pad document numbers for granted patents to 8 digits

with leading zeros when accessing USPTO for PDF documents

CHANGES.rst

@@ -11,6 +11,8 @@ Development
 - [mw] Fix tests
 - [mw] Resolve issue when European publication server returns
   reference to WIPO as HTML response instead of PDF document
+- [mw] Pad document numbers for granted patents to 8 digits
+  with leading zeros when accessing USPTO for PDF documents
 
 
 2019-02-21 0.163.0

patzilla/access/uspto/pdf.py

@@ -54,11 +54,20 @@ def pdf_url(document_number):
     Application: US2016101909A1
     http://pdfaiw.uspto.gov/fdd/09/2016/19/010/0.pdf
 
+    >>> pdf_url('US2016101909A1')
+    'http://pdfaiw.uspto.gov/fdd/09/2016/19/010/0.pdf'
+
     Grant: US10194689B2
     http://pdfpiw.uspto.gov/fdd/89/946/101/0.pdf
+    >>> pdf_url('US10194689B2')
+    'http://pdfpiw.uspto.gov/fdd/89/946/101/0.pdf'
+
+    >>> pdf_url('US2548918')
+    'http://pdfpiw.uspto.gov/fdd/18/489/025/0.pdf'
+
     """
 
-    document = normalize_patent(document_number, for_ops=False, as_dict=True)
+    document = normalize_patent(document_number, for_ops=False, as_dict=True, provider='uspto')
     if not document:
         return
 

patzilla/util/numbers/common.py

@@ -12,11 +12,16 @@
 
 log = logging.getLogger(__name__)
 
+
 class DocumentIdentifierBunch(SmartBunch):
 
     def __str__(self):
         return self.dump()
 
+    def serialize(self):
+        return join_patent(self)
+
+
 def join_patent(patent):
     if not patent:
         return

patzilla/util/numbers/normalize.py

@@ -2,16 +2,20 @@
 # (c) 2007-2018 Andreas Motl <andreas.motl@ip-tools.org>
 import re
 import types
+import logging
 from copy import copy
 from patzilla.util.numbers.denormalize import denormalize_patent_wo
 from patzilla.util.numbers.helper import pad_left, trim_leading_zeros, fullyear_from_year
-from patzilla.util.numbers.common import split_patent_number, join_patent
+from patzilla.util.numbers.common import decode_patent_number, split_patent_number, join_patent
+
+
+logger = logging.getLogger(__name__)
+
 
 """
 Normalize patent- and document-numbers.
 """
 
-
 def patch_patent(patent, provider=None):
 
     if not patent:
@@ -388,6 +392,12 @@ def normalize_patent_wo_pct(patent):
 
 
 def normalize_patent_us(patent, provider=None):
+    """
+    # TODO:
+    # >>> DocumentIdentifier('US2548918').normalize(provider=DocumentProvider.USPTO).serialize()
+    >>> normalize_patent_us(decode_patent_number('US2548918'), provider='uspto').serialize()
+    'US02548918'
+    """
 
     # USPTO number formats
 
@@ -467,14 +477,17 @@ def normalize_patent_us(patent, provider=None):
             padding = '0' * (11 - length)
             patched['number'] = patched['number'][0:4] + padding + patched['number'][4:]
 
-
     # 2018-04-23: Espacenet changed behavior, handle edge case for
     # USD813591S to yield https://worldwide.espacenet.com/publicationDetails/claims?CC=US&NR=D813591S&KC=S
     if provider == 'espacenet':
         if 'number-type' in patched:
             if patched['number-type'] == 'D' and patched['kind'] == 'S':
                 patched['number'] += patched['kind']
 
+    # 2019-03-05: When going to the USPTO itself, pad sequential number to 8 digits.
+    if provider == 'uspto':
+        patched['number'] = patched['number'].zfill(8)
+
     return patched